Adherence in a randomised controlled trial comparing liberal and restrictive red blood cell (RBC) transfusion protocols after cardiac surgery (TITRe2)

Katie Pike,Gavin J Murphy,Chris A Rogers,Rachel Brierley,Barney C Reeves

doi:10.1186/1745-6215-12-s1-a131

Abstract

Objectives The TITRe2 trial is comparing two haemoglobin (Hb) thresholds for RBC transfusion after cardiac surgery, Hb<9.0g/dl (liberal) vs. Hb<7.5g/dl (restrictive). Based on historic data, and with complete adherence, transfusion rates should be 100% in the liberal and 30% in the restrictive group. Convergence of these rates due to non-adherence severely threatens the power of the trial; there is also concern about differential non-adherence, with transfusion being delayed or withheld in the liberal group when Hb remains close to the 9.0g/dl threshold.

Highlights

With complete adherence, transfusion rates should be 100% in the liberal and 30% in the restrictive group. Convergence of these rates due to non-adherence severely threatens the power of the trial; there is concern about differential non-adherence, with transfusion being delayed or withheld in the liberal group when Hb remains close to the 9.0g/dl threshold
56% of participants are being randomised; about 8% of the remaining 44% consented participants breach the 9.0g/dl threshold but are not randomised. 3% of randomised participants are randomised >24 hours after first breaching, but none has been randomised without or before breaching the 9.0g/dl threshold. 32% of participants have had ≥1 instance of non-adherence to the transfusion protocol; in 6%, non-adherence was judged severe
Rates of transfusion in the liberal and restrictive groups are confidential to the Data Monitoring and Ethics Committee (DMEC)